Select Page

# Predicting appendicitis using machine learning in Mathematica

I note more and more published papers on machine learning.  As a clinician, I find it a fascinating way of looking at patient data.  In case you are not familiar with machine learning, the definition given over at Wikipedia is: Machine learning is the subfield of computer science that gives computers the ability to learn without being explicitly programmed. …machine learning explores the study and construction of algorithms that can learn from and make predictions on data.

That is exactly what machine learning is used for in medicine as well.  In a particular branch of machine learning, called supervised learning, a dataset of predictor variables together with a known outcome variable can be passed to the machine, which in turns constructs a model from the data.  A selection of the data is usually kept separately and is used to test the model.  Given that the outcomes are know, it is trivial to calculate the accuracy of the model.  Once a model is generated, data without a known outcome can be passed to the model, which will predict the outcome.  This can indeed be very useful in medicine.

There are many tools available to do machine learning.  I use both Python and Mathematica.  It is really easy to do.  I have put together a short video on YouTube for those familiar with Mathematica, just to show how easy it is.

In the video I use random forest, logistic regression, and support vector machines models to predict the presence of appendicitis from the simulated modified Alvarado score predictor variables.

# Teaching statistics and data science in medical school

Understanding statistical analysis and interpreting the results of research papers are just as important as the ability to correctly diagnose the cause of acute abdominal pain.

Medical knowledge is expanding at a rapid pace. This is evident by the number of research papers being published every year. Although medical students and residents attend a formal education program, it is journal papers that serve as masters of education for the majority of a professional’s life.

The ability to understand the results section of a paper is crucial in deciding to change clinical practice. In order to do this effectively, knowledge of statistics is vital.

Yet, formal training is statistics takes a back seat when it comes to anatomy, physiology, and, clinical teaching. When statitics is part of the curriculum, it is often positioned as less important. It gets even worse when taught with mathematical emphasis. Whilst it may be rigorous to teach using equations, a subset of medical students are lost in this effort.

No medical school can look the other way. Data analysis and computational thinking is part of the future of healthcare. I was reminded of this when I came across this article again, after reading it almost two years ago: NYU medical students learning to analyze big data.

Our efforts at University of Cape Town are growing too. The massive open online course: Understanding clinical research on the Coursera platform, has now had more than 23,000 participants. In the division of General Surgery, I teach the use of data analysis and computational thinking to great effect, using IBM SPSS, Python, Julia, and Mathematica.

It’s time data science and statistical analysis to take its rightful place in medical school curricula.

# The creation of online teaching material as a crisis solution

In an effort to complete the 2016 academic year, the University of Cape Town leadership have called upon the body of lecturers to make use of online and blended teaching material.  The University, as others in the country, are reopening their doors under difficult circumstances.  These relate to continued protest action and the absence of consensus amongst students and staff on the if-and-how of reopening the University.  With classroom attendance expected to be poor or even unwarranted, the problem of providing didactic learning had to be addressed.  The solution, online learning.  A simple call to put recordings of lectures online and to incorporate already existing web-based material.

I am well familiar with this concept.  With more than 1,000 lectures on YouTube, two courses on the massive open online course (MOOC) platform Coursera® (here & here), and an international award in open education from the Open Education Consortium, I am sold on the concept of freeing knowledge from its academic confines.  Knowledge through education is power.  The access to it is a fundamental right and it should not be a commodity.  There can be no better tool to uplift a population, than through proper education.

So now, UCT wants to embrace online education as an instant solution to save the academic year.  So why, after pouring so much energy into the creation of online educational resources, am I not elated, ecstatic, vindicated?  To be honest, I do experience these feelings.  It is, however, mixed with feelings of trepidation, anxiety, and even frustration.

Frustrated, because my plea for the large scale creation of online resources have fallen on deaf ears.  We need only look at the efforts of leading Universities such as the Massachusetts Institute of Technology, Stanford, Harvard and many others that have embraced the online space in their educational efforts.  Not only to the benefit of their local students, but the world at large.  UCT should have been creating these resources at scale a long time ago.

We have to take cognizance of the fact that the efforts of leading Universities took years to develop.  Built with the input of experienced staff and stakeholders.  Experts who know that simply transforming face-to-face teaching or printed material into video and electronic format does not constitute education.  The problem cannot be solved with a purely cognitivist approach and most certainly, not overnight.

There are many problems inherent in the call for the rapid production of online course material.   One glaring example is the lack of formative and summative assessment.  The face-to-face method of providing learning material (lectures), asking a few unstructured questions during lecturing and sitting back in judgement during tests and exams is already a suboptimal approach to education.  When replacing this flawed concept with unstructured online teaching, the outcome must certainly be viewed with concern.  To develop a proper educational resource takes time, effort, experience, research, and most importantly, engagement and consultation with students.  Watch this video from smaccDUB on how students can choreograph their own education.

The call to make online resources available must be supported.  We need to do so in a measured and structured manner, though.  To the University’s credit the Dean of the Health Sciences Faculty has called for the creation of a technology in education committee.  The Centre for Innovation in Learning and Teaching have published an excellent guide to the creation of online educational resources.  Furthermore, they provide individual consultations and hold regular workshops.  Hopefully we can use this opportunity to align our efforts with those of the leading Universities in the world.

# Julia for scientific computing, my second Coursera MOOC

October 2016 has seen the launch of my second course on the Coursera massive open online course (MOOC) platform.  Whereas my first course dealt with the statistics used in healthcare research, this one teaches the new Julia language for scientific computing.  You can find it here.

As with other Coursera offerings, you can pay a nominal fee to get a verified certificate from the University of Cape Town, else you can audit the course for free.  Remember, though, that it is always possible to state that you do not have the financial resources to pay for the verified certificate and Coursera will waive the fee and you will still get your certificate.